Each compositeType consistes of a few components. Use ("bob", 1982) as an example, it contains two components, I assume it is (utf8, integer). So when you want to use a slice query, you need the start and end columns by add components to them. That is what start.addCompo.... and end.addCompo... means. For more information please refer to https://issues.apache.org/jira/browse/CASSANDRA-2231

For you question, I think replace "abc" with "bob" should do the trick.

On Wed, Aug 17, 2011 at 4:56 AM, Alvin UW <alvinuw@gmail.com> wrote:
Hello,

can anyone give an explanation of
start.addComponent("abc", StringSerializer.get()) ;
end.addComponent("abc", StringSerializer.get(), "UTF8Type", AbstractComposite.ComponentEquality.GREATER_THAN_EQUAL) ;


Suppose my composite column names are like ("bob", 1982),  ("bob", 1976). There are more than one "bob" in my table.
now I only know I want to query "bob", but don't know the birth year, how can I get all the column values of "Bob"? with the help of composite type.




2011/8/4 Boris Yen <yulinyen@gmail.com>
Assume you have a column family named "testCF" with comparator CompositeType(AsciiType, IntegerType(reversed=true), IntegerType); and a few columns have been inserted into record key "a".

____________________________________
Composite start = new Composite() ;
Composite end = new Composite() ;
start.addComponent("abc", StringSerializer.get()) ;
end.addComponent("abc", StringSerializer.get(), "UTF8Type", AbstractComposite.ComponentEquality.GREATER_THAN_EQUAL) ;

SliceQuery<String, ByteBuffer, ByteBuffer> query = HFactory.createSliceQuery(keyspace, StringSerializer.get(), ByteBufferSerializer.get(), ByteBufferSerializer.get());
query.setKey("a");
query.setColumnFamily("testCF");
query.setRange(start.serialize(), end.serialize(), false, 100);

List<HColumn<ByteBuffer, ByteBuffer>> columns = query.execute().get().getColumns();


On Thu, Aug 4, 2011 at 3:19 PM, CASSANDRA learner <cassandralearner@gmail.com> wrote:
Can you please gimme an example on this using hector client


On Thu, Aug 4, 2011 at 7:18 AM, Boris Yen <yulinyen@gmail.com> wrote:
It seems to me that your column name consists of two components. If you have the luxury to upgrade your cassandra to 0.8.1+, I think you can think about using the composite type/column. Conceptually, it might suit your use case better.


On Wed, Aug 3, 2011 at 5:28 AM, Eldad Yamin <eldad87@gmail.com> wrote:
Hello,
I wonder if I can select a column or all columns that start with X.
E.g I have columns ABC_1, ABC_2, ZZZ_1 and I want to select all columns that start with ABC_ - is that possible?



Thanks!